Experiment

Stimuli

  • Stimuli were selected from the previous study (the sound of teaching music/experiment-2).
  • 36 stimuli were randomly chosen from each technique (either articulation or dynamics - therefore in total 72 stimuli). The half of them was selected from the teaching condition and the other half was from the performing condition.
    • 18 for teaching-articulation performances
    • 18 for performing-articulation performances
    • 18 for teaching-dynamics performances
    • 18 for performing-dynamics performances

Design

  • 2 blocks (articulation / dynamics): each block only contained performances with either of articulation or dynamics. The order of the blocks was counterbalanced across participants.
  • 36 trials for each block (each stimulus was presented only once) and the stimuli were randomly presented within the block.

Procedure

  1. Musicians (> 6yo experience) were asked to listen to a number of recordings and judge whether each performance was produced for teaching purposes or not.

Actual instruction:

Each performance was produced in order to either 1) teach the musical expressive technique (as a teacher) or 2) perform their best (as a performer).

You will be asked to judge whether each performer had the intention to teach or not by pressing the 'Yes' <Left> or 'No' <Right> key.
  1. Participants pressed “YES” if they think the recording was produced for teaching (as a teacher) whereas they pressed “NO” if they thought the recording was produced for performing (as a performer).

Variables

  • IOIs (tempo)
  • IOIs at transition points (tempo)
  • CV (tempo)
  • KOT (articulation)
  • KV (dynamics)
  • KV-Diff (dynamics)

Judged as “teaching” (%) for each stimulus

Results

1. Correlations

IOIs

All

## `geom_smooth()` using formula 'y ~ x'

Articulation (n.s.)

## 
##  Shapiro-Wilk normality test
## 
## data:  ioi[Skill == "articulation"]$Teaching
## W = 0.97385, p-value = 0.5398

## 
##  Shapiro-Wilk normality test
## 
## data:  ioi[Skill == "articulation"]$Mean
## W = 0.94184, p-value = 0.05796
## 
##  Pearson's product-moment correlation
## 
## data:  ioi[Skill == "articulation"]$Teaching and ioi[Skill == "articulation"]$Mean
## t = 1.4838, df = 34, p-value = 0.1471
## alternative hypothesis: true correlation is not equal to 0
## 95 percent confidence interval:
##  -0.08915582  0.53203453
## sample estimates:
##       cor 
## 0.2466031

Dynamics (n.s.)

## 
##  Shapiro-Wilk normality test
## 
## data:  ioi[Skill == "dynamics"]$Teaching
## W = 0.93654, p-value = 0.03968

## 
##  Shapiro-Wilk normality test
## 
## data:  ioi[Skill == "dynamics"]$Mean
## W = 0.94622, p-value = 0.07951
## 
##  Pearson's product-moment correlation
## 
## data:  ioi[Skill == "dynamics"]$Teaching and ioi[Skill == "dynamics"]$Mean
## t = 2.4731, df = 34, p-value = 0.01855
## alternative hypothesis: true correlation is not equal to 0
## 95 percent confidence interval:
##  0.07104344 0.63725327
## sample estimates:
##       cor 
## 0.3904656
## 
##  Spearman's rank correlation rho
## 
## data:  ioi[Skill == "dynamics"]$Teaching and ioi[Skill == "dynamics"]$Mean
## S = 5549.6, p-value = 0.09111
## alternative hypothesis: true rho is not equal to 0
## sample estimates:
##       rho 
## 0.2857712

Tempo at transition points

All

## `geom_smooth()` using formula 'y ~ x'

Articulation (n.s.)

## 
##  Shapiro-Wilk normality test
## 
## data:  ioi_tra[Skill == "articulation"]$Teaching
## W = 0.97385, p-value = 0.5398

## 
##  Shapiro-Wilk normality test
## 
## data:  ioi_tra[Skill == "articulation"]$Mean
## W = 0.86402, p-value = 0.000407
## 
##  Pearson's product-moment correlation
## 
## data:  ioi_tra[Skill == "articulation"]$Teaching and ioi_tra[Skill == "articulation"]$Mean
## t = -0.21462, df = 34, p-value = 0.8313
## alternative hypothesis: true correlation is not equal to 0
## 95 percent confidence interval:
##  -0.3609560  0.2953224
## sample estimates:
##        cor 
## -0.0367821
## 
##  Spearman's rank correlation rho
## 
## data:  ioi_tra[Skill == "articulation"]$Teaching and ioi_tra[Skill == "articulation"]$Mean
## S = 7366, p-value = 0.7633
## alternative hypothesis: true rho is not equal to 0
## sample estimates:
##        rho 
## 0.05198897

Dynamics (*)

## 
##  Shapiro-Wilk normality test
## 
## data:  ioi_tra[Skill == "dynamics"]$Teaching
## W = 0.93654, p-value = 0.03968

## 
##  Shapiro-Wilk normality test
## 
## data:  ioi_tra[Skill == "dynamics"]$Mean
## W = 0.68463, p-value = 1.637e-07
## 
##  Pearson's product-moment correlation
## 
## data:  ioi_tra[Skill == "dynamics"]$Teaching and ioi_tra[Skill == "dynamics"]$Mean
## t = 3.0482, df = 34, p-value = 0.004434
## alternative hypothesis: true correlation is not equal to 0
## 95 percent confidence interval:
##  0.1589380 0.6872201
## sample estimates:
##       cor 
## 0.4632826
## 
##  Spearman's rank correlation rho
## 
## data:  ioi_tra[Skill == "dynamics"]$Teaching and ioi_tra[Skill == "dynamics"]$Mean
## S = 4873.1, p-value = 0.02513
## alternative hypothesis: true rho is not equal to 0
## sample estimates:
##       rho 
## 0.3728308

CV (tempo variability)

All

## `geom_smooth()` using formula 'y ~ x'

Articulation (n.s.)

## 
##  Shapiro-Wilk normality test
## 
## data:  cv[Skill == "articulation"]$Teaching
## W = 0.97385, p-value = 0.5398

## 
##  Shapiro-Wilk normality test
## 
## data:  cv[Skill == "articulation"]$CV
## W = 0.72145, p-value = 6.255e-07
## 
##  Pearson's product-moment correlation
## 
## data:  cv[Skill == "articulation"]$Teaching and cv[Skill == "articulation"]$CV
## t = -1.1432, df = 34, p-value = 0.261
## alternative hypothesis: true correlation is not equal to 0
## 95 percent confidence interval:
##  -0.4899546  0.1453349
## sample estimates:
##        cor 
## -0.1923869
## 
##  Spearman's rank correlation rho
## 
## data:  cv[Skill == "articulation"]$Teaching and cv[Skill == "articulation"]$CV
## S = 8455.3, p-value = 0.609
## alternative hypothesis: true rho is not equal to 0
## sample estimates:
##        rho 
## -0.0882002

Dynamics (n.s.)

## 
##  Shapiro-Wilk normality test
## 
## data:  cv[Skill == "dynamics"]$Teaching
## W = 0.93654, p-value = 0.03968

## 
##  Shapiro-Wilk normality test
## 
## data:  cv[Skill == "dynamics"]$CV
## W = 0.68893, p-value = 1.904e-07
## 
##  Pearson's product-moment correlation
## 
## data:  cv[Skill == "dynamics"]$Teaching and cv[Skill == "dynamics"]$CV
## t = 1.3111, df = 34, p-value = 0.1986
## alternative hypothesis: true correlation is not equal to 0
## 95 percent confidence interval:
##  -0.1176402  0.5110755
## sample estimates:
##       cor 
## 0.2193741
## 
##  Spearman's rank correlation rho
## 
## data:  cv[Skill == "dynamics"]$Teaching and cv[Skill == "dynamics"]$CV
## S = 6261.9, p-value = 0.2567
## alternative hypothesis: true rho is not equal to 0
## sample estimates:
##       rho 
## 0.1940868

KOT

All

## `geom_smooth()` using formula 'y ~ x'

Legato (n.s.)

## 
##  Shapiro-Wilk normality test
## 
## data:  kot_all[Subcomponent == "Legato"]$Teaching
## W = 0.97385, p-value = 0.5398

## 
##  Shapiro-Wilk normality test
## 
## data:  kot_all[Subcomponent == "Legato"]$Mean
## W = 0.97859, p-value = 0.6976
## 
##  Pearson's product-moment correlation
## 
## data:  kot_all[Subcomponent == "Legato"]$Teaching and kot_all[Subcomponent == "Legato"]$Mean
## t = -0.15232, df = 34, p-value = 0.8798
## alternative hypothesis: true correlation is not equal to 0
## 95 percent confidence interval:
##  -0.3516330  0.3050387
## sample estimates:
##         cor 
## -0.02611412

Staccato (n.s.)

## 
##  Shapiro-Wilk normality test
## 
## data:  kot_all[Subcomponent == "Staccato"]$Teaching
## W = 0.97385, p-value = 0.5398

## 
##  Shapiro-Wilk normality test
## 
## data:  kot_all[Subcomponent == "Staccato"]$Mean
## W = 0.96129, p-value = 0.2355
## 
##  Pearson's product-moment correlation
## 
## data:  kot_all[Subcomponent == "Staccato"]$Teaching and kot_all[Subcomponent == "Staccato"]$Mean
## t = -0.87009, df = 34, p-value = 0.3904
## alternative hypothesis: true correlation is not equal to 0
## 95 percent confidence interval:
##  -0.4541034  0.1901707
## sample estimates:
##        cor 
## -0.1475859

Forte (n.s.)

## 
##  Shapiro-Wilk normality test
## 
## data:  kot_all[Subcomponent == "Forte"]$Teaching
## W = 0.93654, p-value = 0.03968

## 
##  Shapiro-Wilk normality test
## 
## data:  kot_all[Subcomponent == "Forte"]$Mean
## W = 0.78286, p-value = 7.57e-06
## 
##  Pearson's product-moment correlation
## 
## data:  kot_all[Subcomponent == "Forte"]$Teaching and kot_all[Subcomponent == "Forte"]$Mean
## t = -0.64948, df = 34, p-value = 0.5204
## alternative hypothesis: true correlation is not equal to 0
## 95 percent confidence interval:
##  -0.4238216  0.2260574
## sample estimates:
##        cor 
## -0.1106998
## 
##  Spearman's rank correlation rho
## 
## data:  kot_all[Subcomponent == "Forte"]$Teaching and kot_all[Subcomponent == "Forte"]$Mean
## S = 9115.5, p-value = 0.3125
## alternative hypothesis: true rho is not equal to 0
## sample estimates:
##        rho 
## -0.1731673

Piano (*)

## 
##  Shapiro-Wilk normality test
## 
## data:  kot_all[Subcomponent == "Piano"]$Teaching
## W = 0.93654, p-value = 0.03968

## 
##  Shapiro-Wilk normality test
## 
## data:  kot_all[Subcomponent == "Piano"]$Mean
## W = 0.89598, p-value = 0.002653
## 
##  Pearson's product-moment correlation
## 
## data:  kot_all[Subcomponent == "Piano"]$Teaching and kot_all[Subcomponent == "Piano"]$Mean
## t = -2.2047, df = 34, p-value = 0.03434
## alternative hypothesis: true correlation is not equal to 0
## 95 percent confidence interval:
##  -0.61118593 -0.02843273
## sample estimates:
##        cor 
## -0.3536648
## 
##  Spearman's rank correlation rho
## 
## data:  kot_all[Subcomponent == "Piano"]$Teaching and kot_all[Subcomponent == "Piano"]$Mean
## S = 10597, p-value = 0.02913
## alternative hypothesis: true rho is not equal to 0
## sample estimates:
##        rho 
## -0.3638967

KV

All

## `geom_smooth()` using formula 'y ~ x'

Forte (**)

## 
##  Shapiro-Wilk normality test
## 
## data:  vel_all[Subcomponent == "Forte"]$Teaching
## W = 0.93654, p-value = 0.03968

## 
##  Shapiro-Wilk normality test
## 
## data:  vel_all[Subcomponent == "Forte"]$Mean
## W = 0.93909, p-value = 0.04761
## 
##  Pearson's product-moment correlation
## 
## data:  vel_all[Subcomponent == "Forte"]$Teaching and vel_all[Subcomponent == "Forte"]$Mean
## t = 2.8987, df = 34, p-value = 0.006517
## alternative hypothesis: true correlation is not equal to 0
## 95 percent confidence interval:
##  0.1365985 0.6749772
## sample estimates:
##       cor 
## 0.4451568
## 
##  Spearman's rank correlation rho
## 
## data:  vel_all[Subcomponent == "Forte"]$Teaching and vel_all[Subcomponent == "Forte"]$Mean
## S = 4460.3, p-value = 0.009591
## alternative hypothesis: true rho is not equal to 0
## sample estimates:
##      rho 
## 0.425953

Piano (*)

## 
##  Shapiro-Wilk normality test
## 
## data:  vel_all[Subcomponent == "Piano"]$Teaching
## W = 0.93654, p-value = 0.03968

## 
##  Shapiro-Wilk normality test
## 
## data:  vel_all[Subcomponent == "Piano"]$Mean
## W = 0.90118, p-value = 0.003677
## 
##  Pearson's product-moment correlation
## 
## data:  vel_all[Subcomponent == "Piano"]$Teaching and vel_all[Subcomponent == "Piano"]$Mean
## t = -2.9072, df = 34, p-value = 0.006378
## alternative hypothesis: true correlation is not equal to 0
## 95 percent confidence interval:
##  -0.6756828 -0.1378714
## sample estimates:
##        cor 
## -0.4461963
## 
##  Spearman's rank correlation rho
## 
## data:  vel_all[Subcomponent == "Piano"]$Teaching and vel_all[Subcomponent == "Piano"]$Mean
## S = 10972, p-value = 0.01252
## alternative hypothesis: true rho is not equal to 0
## sample estimates:
##        rho 
## -0.4120633

Legato (n.s.)

## 
##  Shapiro-Wilk normality test
## 
## data:  vel_all[Subcomponent == "Legato"]$Teaching
## W = 0.97385, p-value = 0.5398

## 
##  Shapiro-Wilk normality test
## 
## data:  vel_all[Subcomponent == "Legato"]$Mean
## W = 0.94958, p-value = 0.1014
## 
##  Pearson's product-moment correlation
## 
## data:  vel_all[Subcomponent == "Legato"]$Teaching and vel_all[Subcomponent == "Legato"]$Mean
## t = 0.48723, df = 34, p-value = 0.6292
## alternative hypothesis: true correlation is not equal to 0
## 95 percent confidence interval:
##  -0.2521651  0.4008392
## sample estimates:
##        cor 
## 0.08326912

Staccato (n.s.)

## 
##  Shapiro-Wilk normality test
## 
## data:  vel_all[Subcomponent == "Staccato"]$Teaching
## W = 0.97385, p-value = 0.5398

## 
##  Shapiro-Wilk normality test
## 
## data:  vel_all[Subcomponent == "Staccato"]$Mean
## W = 0.9757, p-value = 0.5997
## 
##  Pearson's product-moment correlation
## 
## data:  vel_all[Subcomponent == "Staccato"]$Teaching and vel_all[Subcomponent == "Staccato"]$Mean
## t = 1.247, df = 34, p-value = 0.2209
## alternative hypothesis: true correlation is not equal to 0
## 95 percent confidence interval:
##  -0.1282152  0.5030993
## sample estimates:
##       cor 
## 0.2091298

KV-Diff

All

## `geom_smooth()` using formula 'y ~ x'

Forte to Piano (***)

## 
##  Shapiro-Wilk normality test
## 
## data:  vel_diff_all[Subcomponent == "FtoP"]$Teaching
## W = 0.93654, p-value = 0.03968

## 
##  Shapiro-Wilk normality test
## 
## data:  vel_diff_all[Subcomponent == "FtoP"]$Mean
## W = 0.98328, p-value = 0.8496
## 
##  Pearson's product-moment correlation
## 
## data:  vel_diff_all[Subcomponent == "FtoP"]$Teaching and vel_diff_all[Subcomponent == "FtoP"]$Mean
## t = -6.5455, df = 34, p-value = 1.7e-07
## alternative hypothesis: true correlation is not equal to 0
## 95 percent confidence interval:
##  -0.8634165 -0.5540757
## sample estimates:
##        cor 
## -0.7466889
## 
##  Spearman's rank correlation rho
## 
## data:  vel_diff_all[Subcomponent == "FtoP"]$Teaching and vel_diff_all[Subcomponent == "FtoP"]$Mean
## S = 13407, p-value = 5.563e-07
## alternative hypothesis: true rho is not equal to 0
## sample estimates:
##        rho 
## -0.7254799

Piano to Forte (***)

## 
##  Shapiro-Wilk normality test
## 
## data:  vel_diff_all[Subcomponent == "PtoF"]$Teaching
## W = 0.93654, p-value = 0.03968

## 
##  Shapiro-Wilk normality test
## 
## data:  vel_diff_all[Subcomponent == "PtoF"]$Mean
## W = 0.91318, p-value = 0.00798
## 
##  Pearson's product-moment correlation
## 
## data:  vel_diff_all[Subcomponent == "PtoF"]$Teaching and vel_diff_all[Subcomponent == "PtoF"]$Mean
## t = 4.2234, df = 34, p-value = 0.0001699
## alternative hypothesis: true correlation is not equal to 0
## 95 percent confidence interval:
##  0.3196656 0.7672660
## sample estimates:
##       cor 
## 0.5865961
## 
##  Spearman's rank correlation rho
## 
## data:  vel_diff_all[Subcomponent == "PtoF"]$Teaching and vel_diff_all[Subcomponent == "PtoF"]$Mean
## S = 3340.9, p-value = 0.0002843
## alternative hypothesis: true rho is not equal to 0
## sample estimates:
##       rho 
## 0.5700193

Legato to Staccato (n.s.)

## 
##  Shapiro-Wilk normality test
## 
## data:  vel_diff_all[Subcomponent == "LtoS"]$Teaching
## W = 0.97385, p-value = 0.5398

## 
##  Shapiro-Wilk normality test
## 
## data:  vel_diff_all[Subcomponent == "LtoS"]$Mean
## W = 0.94562, p-value = 0.07613
## 
##  Pearson's product-moment correlation
## 
## data:  vel_diff_all[Subcomponent == "LtoS"]$Teaching and vel_diff_all[Subcomponent == "LtoS"]$Mean
## t = 1.3835, df = 34, p-value = 0.1755
## alternative hypothesis: true correlation is not equal to 0
## 95 percent confidence interval:
##  -0.105688  0.519962
## sample estimates:
##       cor 
## 0.2308638

Staccato to Legato (*)

## 
##  Shapiro-Wilk normality test
## 
## data:  vel_diff_all[Subcomponent == "StoL"]$Teaching
## W = 0.97385, p-value = 0.5398

## 
##  Shapiro-Wilk normality test
## 
## data:  vel_diff_all[Subcomponent == "StoL"]$Mean
## W = 0.95266, p-value = 0.1268
## 
##  Pearson's product-moment correlation
## 
## data:  vel_diff_all[Subcomponent == "StoL"]$Teaching and vel_diff_all[Subcomponent == "StoL"]$Mean
## t = 2.2252, df = 34, p-value = 0.03281
## alternative hypothesis: true correlation is not equal to 0
## 95 percent confidence interval:
##  0.03171544 0.61324012
## sample estimates:
##       cor 
## 0.3565362

2. Partical correlation (only for KOT, KV, KV-Diff)

KOT

Legato (n.s.)

Estimate

  • partial correlation coefficient between two variables (Teaching vs KOT)

pcor.test(partial[Subcomponent == "Legato"]$KOT, partial[Subcomponent == "Legato"]$Teaching, partial[Subcomponent == "Legato", c("IOI", "KV", "KVDiff")])

Staccato (n.s.)

Estimate

  • partial correlation coefficient between two variables (Teaching vs KOT)

pcor.test(partial[Subcomponent == "Staccato"]$KOT, partial[Subcomponent == "Staccato"]$Teaching, partial[Subcomponent == "Staccato", c("IOI", "KV", "KVDiff")])

KV

Forte (n.s.)

Estimate

  • partial correlation coefficient between two variables (Teaching vs KV)

pcor.test(partial[Subcomponent == "Forte"]$KV, partial[Subcomponent == "Forte"]$Teaching, partial[Subcomponent == "Forte", c("IOI", "KOT", "KVDiff")])

Piano (n.s.)

Estimate

  • partial correlation coefficient between two variables (Teaching vs KV)

pcor.test(partial[Subcomponent == "Piano"]$KV, partial[Subcomponent == "Piano"]$Teaching, partial[Subcomponent == "Piano", c("IOI", "KOT", "KVDiff")])

KV-Diff

Forte to Piano (***)

Estimate

  • partial correlation coefficient between two variables (Teaching vs KVDiff)

pcor.test(partial[Subcomponent2 == "FtoP"]$KVDiff, partial[Subcomponent2 == "FtoP"]$Teaching, partial[Subcomponent2 == "FtoP", c("IOI", "KOT", "KV")])

Piano to Forte (***)

Estimate

  • partial correlation coefficient between two variables (Teaching vs KVDiff)

pcor.test(partial[Subcomponent2 == "PtoF"]$KVDiff, partial[Subcomponent2 == "PtoF"]$Teaching, partial[Subcomponent2 == "PtoF", c("IOI", "KOT", "KV")])

Multiple regression

Note: additive - no interaction considered

Legato

m1 <- lm(Teaching ~ IOI + KOT + KV + KVDiff, data = partial[Subcomponent == "Legato"])
summary(m1)
## 
## Call:
## lm(formula = Teaching ~ IOI + KOT + KV + KVDiff, data = partial[Subcomponent == 
##     "Legato"])
## 
## Residuals:
##     Min      1Q  Median      3Q     Max 
## -29.722 -10.706   1.336   9.958  24.511 
## 
## Coefficients:
##               Estimate Std. Error t value Pr(>|t|)   
## (Intercept) -184.31750   75.50965  -2.441  0.02056 * 
## IOI            0.58277    0.21572   2.702  0.01109 * 
## KOT           -0.08497    0.11520  -0.738  0.46631   
## KV             1.10473    0.44163   2.501  0.01786 * 
## KVDiff         1.89934    0.64740   2.934  0.00625 **
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 14.89 on 31 degrees of freedom
## Multiple R-squared:  0.2894, Adjusted R-squared:  0.1977 
## F-statistic: 3.156 on 4 and 31 DF,  p-value: 0.02752
check_model(m1)

Staccato

m2 <- lm(Teaching ~ IOI + KOT + KV + KVDiff, data = partial[Subcomponent == "Staccato"])
summary(m2)
## 
## Call:
## lm(formula = Teaching ~ IOI + KOT + KV + KVDiff, data = partial[Subcomponent == 
##     "Staccato"])
## 
## Residuals:
##     Min      1Q  Median      3Q     Max 
## -27.617  -9.232   1.271   9.239  27.628 
## 
## Coefficients:
##             Estimate Std. Error t value Pr(>|t|)  
## (Intercept) -65.8795    70.5859  -0.933   0.3579  
## IOI           0.1210     0.2993   0.404   0.6889  
## KOT          -0.1780     0.2030  -0.877   0.3873  
## KV            0.5797     0.3986   1.454   0.1559  
## KVDiff        0.7302     0.3755   1.945   0.0609 .
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 15.57 on 31 degrees of freedom
## Multiple R-squared:  0.2231, Adjusted R-squared:  0.1228 
## F-statistic: 2.225 on 4 and 31 DF,  p-value: 0.08915
check_model(m2)

Forte

m3 <- lm(Teaching ~ IOI + KOT + KV + KVDiff, data = partial[Subcomponent == "Forte"])
summary(m3)
## 
## Call:
## lm(formula = Teaching ~ IOI + KOT + KV + KVDiff, data = partial[Subcomponent == 
##     "Forte"])
## 
## Residuals:
##     Min      1Q  Median      3Q     Max 
## -20.844  -7.451  -1.172   8.217  26.761 
## 
## Coefficients:
##              Estimate Std. Error t value Pr(>|t|)    
## (Intercept) -60.40467   80.97584  -0.746 0.461312    
## IOI           0.28747    0.22992   1.250 0.220532    
## KOT          -0.10305    0.06189  -1.665 0.105986    
## KV            0.02354    0.40267   0.058 0.953766    
## KVDiff       -1.63853    0.38204  -4.289 0.000163 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 12.69 on 31 degrees of freedom
## Multiple R-squared:  0.6518, Adjusted R-squared:  0.6069 
## F-statistic: 14.51 on 4 and 31 DF,  p-value: 8.785e-07
check_model(m3)

Piano

m4 <- lm(Teaching ~ IOI + KOT + KV + KVDiff, data = partial[Subcomponent == "Piano"])
summary(m4)
## 
## Call:
## lm(formula = Teaching ~ IOI + KOT + KV + KVDiff, data = partial[Subcomponent == 
##     "Piano"])
## 
## Residuals:
##     Min      1Q  Median      3Q     Max 
## -24.022  -9.992  -1.424   9.291  24.718 
## 
## Coefficients:
##               Estimate Std. Error t value Pr(>|t|)    
## (Intercept) -106.08277   89.12100  -1.190 0.242955    
## IOI            0.55404    0.22918   2.417 0.021700 *  
## KOT           -0.06475    0.09523  -0.680 0.501636    
## KV            -0.65679    0.63689  -1.031 0.310404    
## KVDiff         1.09161    0.24538   4.449 0.000104 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 14.38 on 31 degrees of freedom
## Multiple R-squared:  0.5526, Adjusted R-squared:  0.4949 
## F-statistic: 9.574 on 4 and 31 DF,  p-value: 3.681e-05
check_model(m4)